...
At the outset the ontology team recognized the existence of a great deal of prior art in the form of published ontologies and significant ongoing ontology initiatives addressing the representation of bibliographic information in RDF. Elements of the Bibliographic Ontology and FaBIO had already been incorporated into the VIVO-ISF Ontology (GitHub) and were familiar to team members from previous work – and Paolo Ciccarese from Harvard was a principal FaBIO contributor. The BIBFRAME initiative at the Library of Congress addresses the representation of MARC metadata in RDF, while OCLC has worked to extend the Schema.org ontology as a bridge between the library community and the Web. The Collections ontology and ORE address collections of digital objects; the Open Annotation Data Model annotations, and PROV-O and PAV provenance.
| Panel | ||||||
|---|---|---|---|---|---|---|
| ||||||
SRSIS OntologyBecause no existing ontology supports the range of entities and relationship that SRSIS will encompass, we will use the Protégé ontology editor to develop a SRSIS ontology framework that reuses appropriate parts of currently available ontologies while introducing extensions and additions where necessary. The framework will be based on and remain compatible with the existing VIVO and emerging research dataset and research resource ontology work. It will be sufficiently expressive to encompass traditional catalog metadata from Cornell, Stanford, and Harvard; the basic linked data elements described in the Stanford Linked Data Workshop Technology Plan; and the usage and other contextual elements from StackLife. The ontology will capture a series of basic concepts and be structured as modules that draw inspiration from and reuse existing ontology classes and properties where appropriate, such as the Semantic Publishing and Referencing ontologies, and that also support arbitrary system-wide refinement, including local extensions. |
Ontology Team Activities, 2016
Engaging with BIBFRAME
The LD4L ontology team
...
spent much of 2015 engaged in the evolution of BIBFRAME, namely through the BIBFRAME discussion list and directly with the Office of Network Standards at the Library of Congress. In April Rob Sanderson submitted Analysis of the BIBFRAME Ontology for Linked Data Best Practices to the Library of Congress; the document recommends changes to BIBFRAME to better reflect best practices in the Linked Data domain so as not to marginalize the RDF data libraries contribute to the semantic web.
Following on Rob Sanderson's and LD4L's recommended changes to BIBFRAME, the team wrote a bibliographic ontology with the hope that derivations from BIBFRAME found in this ontology will be folded into the official BIBFRAME namespace. The proposed changes include, but are not limited to, incorporation of the following conventions:
Reuse stable pre-existing classes and properties from external ontologies rather than declaring new ones within the BIBFRAME namespace.
Use URIs rather than strings to identify resources.
Replace the bf:Authority classes with Real World Entity classes for people, places, things, etc.
Define only one pattern to model one feature of the knowledge domain.
Clarify the directionality of properties via naming, definitions, and, where applicable, domain and range constraints, and add inverse properties where appropriate.
Name terms consistently, and make the distinction between classes, object properties, and datatype properties clear through standard naming conventions.
Efforts have been made to consider each class and property, but this ontology is largely untested. LD4L may provide revised/expanded versions in future as we identify new use cases, we begin to test the ontology with instance data, and BIBFRAME 2.0 revisions solidify. The RDF generated as an output of the project will be based on the LD4L ontology and will be made available for testing. For future consideration, there is a need to not only align the LD4L ontology with BIBFRAME 2.0, but also Schema.org (we’ve had early conversations with OCLC about this), the Doremus Project (using FRBRoo), Europeana’s Data Model (EDM), and other RDF models within the bibliographic and cultural heritage domains.
Preprocessing Metadata for Richer RDF Conversions
Early experience in the project suggested preprocessing MARC data to include URIs for entities referred to records would significantly increase the value of the RDF output from the BIBFRAME converter. To date, in both the MARC specification and in cataloging practices, the use of URIs to identify entities is uneven. To this end, URIs in MARC: A Call for Best Practices was submitted to the Program for Cooperative Cataloging (PCC) discussion list in May of to spur conversation on the importance of URIs in library data and standardizing how they are stored in MARC. A number of outcomes have occurred out the document:
The MARC Advisory Committee (MAC) discussed the paper with members of LD4L and recommended formal proposals be submitted for areas where MARC could better support the storage of URIs with consistent semantics.
The PCC Task Group on URIs in MARC was formed to address immediate policy issues and develop a work plan with related guidelines in collaboration with MAC, library platform vendors and open source developers, and linked data experts (some from LD4L).
In parallel with the PCC URI Task Group, an number of other organizations (e.g. the British Library) have begun to make MAC proposals related to URIs in MARC.
Because such a large portion of MARC records in libraries are provided by vendors, Nancy Lorimer has created a preliminary set of vendor specifications for enhancing MARC records. The specifications include both recommendations for the addition of URIs and for adjusting cataloging procedures to include specific fields and vocabularies that enhance the transformation into BIBFRAME. Adjustments to these specifications will be ongoing as the BIBFRAME converter is improved and modified to accommodate BIBFRAME 2.0. She and Phil Schreur will also be meeting with Casalini Libri at ALA Midwinter 2016 to discuss implementing these specifications.
Continued Work with Global URIs
In the fall of 2015 Linked Data for Libraries (LD4L) hosted a conversation with members of the CONSER, VIVO, and ISSN communities to discuss global URIs for serials. As the provider of unique identifiers that underpin the systems currently driving much of the continuing resources ecosystem, the ISSN network brings together a wealth of domain expertise and an established network of stakeholders. ISSN is well positioned to have a significant role in providing global URIs and linked data for serials.
The LD4L community is heartened to see the growing interest in modeling serial works as RDF with permanent, stable URIs. ISSN’s current practice of creating identifiers for different serial entities (e.g., ISSN-L, e-ISSN, and p-ISSN, and ISSN for language versions) is very much in parallel with the task of minting URIs for different entity types. These practices provide a sound foundation for further progress. For example, a URI and modeling that explicitly collates all the various versions of a serial would be valuable to connect different languages and formats via a single identifier while preserving the full granularity of current ISSN practice.
We understand that there are number of initiatives underway or scheduled to happen that may have an effect on ISSN’s work in this arena, including the CONSER Operations Committee’s involvement with BIBFRAME and the 2016 review of ISO-3297. We hope there is a convergence on these fronts to ensure the most viable solution for establishing global URIs at the earliest and most authoritative moments in the lifecycle of a serial.
LD4L is particularly happy to hear that publishing URIs and linked data for serials is under consideration at ISSN. We have a strong interest in adopting and promoting ISSN linked data and would endorse further investigation of options for making complete, current, and open RDF descriptions of entities available to data consumers. We are confident that stable identifiers for ISSNs provided in linked data format will encourage widespread adoption of any URIs that ISSN creates. Ideally ISSN would adhere to Linked Data Best Practices, including:
Provide persistent links.
Provide useful data through standard protocols (content negotiation, SPARQL, etc.).
Provide useful links to other entities. ISSN tracks how different serial entities relate to each other and their respective agents; without access to these relationships the data is far less useful.
Provide the data according to model/s that meet data consumer needs; consulting the CONSER BIBFRAME Task Force and experienced linked data practitioners would be advisable for determining the model/s.
If ISSN continues to pursue a linked data agenda, LD4L is ready and willing to contribute to the discussion and testing. We look forward to ISSN’s efforts in this space.
Ontology team activities to 2015
Local vs. global identifiers
...